Lanthanum and Praseodymium Bromide Pnictides. A Convergence of Interstitial Chemistry in Cluster Halides and Intermetallic Pnictides

Lanthanum and Praseodymium Bromide Pnictides. A Convergence of Interstitial Chemistry in Cluster Halides and Intermetallic Pnictides

Journal of Solid State Chemistry 144, 175—180 (1999) Article ID jssc.1999.8143, available online at http://www.idealibrary.com on Lanthanum and Prase...

281KB Sizes 0 Downloads 29 Views

Journal of Solid State Chemistry 144, 175—180 (1999) Article ID jssc.1999.8143, available online at http://www.idealibrary.com on

Lanthanum and Praseodymium Bromide Pnictides. A Convergence of Interstitial Chemistry in Cluster Halides and Intermetallic Pnictides Elizabeth A. Jensen, Laura M. Hoistad, and John D. Corbett Department of Chemistry, Iowa State University, Ames, Iowa 50011 Received August 24, 1998; in revised form December 18, 1998; accepted December 22, 1998

Reactions of metal R, RBr3 (R ⴝ La, Pr), and pnictogen Pn (P, As, Sb, Bi) at 900–1000°C provide a new series of condensed cluster products, either cubic R3Br3Pn or hexagonal R5 Pn3Br, but not both for the same Pn. Syntheses and lattice dimensions for all eight examples are reported. Cubic R3Br3 Pn results with interstitial P or As (Ca3I3P type, I4132 (No. 214), Z ⴝ 8), whereas hexagonal P5 Pn3Br is formed for Pn ⴝ Sb, Bi with interstitial Br (stuffed Mn5Si3 type), a new range of composition for these compounds. The latter structure for La5Sb3Br was refined by single-crystal X-ray diffraction data (P63/mcm (No. 193), Z ⴝ 2, a ⴝ 9.602(2) As , c ⴝ 6.7331(3) As , R(F)/Rw ⴝ 1.7/2.0%). Features that determine the formation of a particular set of compounds in each structure type are discussed.  1999 Academic Press

1. INTRODUCTION

A large quantity of research on rare-earth metal (R) cluster and condensed cluster halides (X) has been carried out in the past 20 years. This has revealed a large number of compounds in which characteristic R X -type clusters   may be either discrete or condensed in one, two, or three dimensions into oligomers or, more often, infinite arrays. Substantially all of these require an interstitial atom (Z) within each cluster unit for stability. The interstitial atoms may be nonmetals, e.g., B, C, N, Si, or one of many transition metals, i.e., Mn—Cu, Ru—Pd, Re—Au. Stoichiometries for binary phases range from RXZ in condensed systems to R X Z with discrete clusters (1, 2). The varieties of   structures expand considerably when quaternary alkalimetal derivatives are included (3). Another large family of discrete cluster halides that requires interstitials was encountered first for zirconium, but virtually no isostructural relatives exist among the rare-earth element halides because of the substantially different stoichiometries required to meet basically the same electronic rules for cluster stability (2, 4). Chemistries in the neighboring chalcogenide cluster systems are only distantly related, e.g., as the Chevrel phases.

One cluster halide and structure type we are interested in presently is that of R Br Pn, a three-dimensional array of   condensed clusters known first for Gd Cl C (5), then for   R I Z, R"La, Pr; Z"Os, Ir, Pt (6) and Pr Br Z, Z"Co,     Os, Rh, Ir, Pt (7). These occur in the cubic Ca I P structure  type, basically a stuffed NaCl derivative (8) and, in contrast, a valence compound. The 3D network is built from Zcentered clusters that each share three nonadjacent edges with other clusters. A widely different family of compounds that shares only the characteristic of a diverse interstitial chemistry has been established and characterized in intermetallic systems. The host binaries all have a common A B stoichiometry and an   Mn Si -type structure, the important part of which is a con  facial A B chain of trigonal antiprisms of A with the   shared faces bridged by B. These are basically all electronrich and metallic. The binary examples are widespread, and for the rare-earth elements they are known for most of the tetrel (Si—Pb) and pnictogen families (9). These are relevant here because apparently all examples share the characteristic of A B Z formation with interstitial Z centered in each   antiprism (9, 10). For example, the ternary La Ge Z (11)   and La Pb Z (12) series are each known with 18 to 20   different Z, the last spanning many of the lighter half of the main-group elements as well as the later 3d elements. Most of these products are still metallic and not electron precise. On the other hand, the known R Pn pnictide hosts are   virtually uninvestigated with respect to their Z chemistry other than R Pn H for R"Eu, Yb (13).   Our synthetic explorations of R—Br—Pn systems (R"La, Pr) for new cluster halide chemistry have led to a surprising confluence of these two very different areas of interstitial chemistry with Br vs Pn interstitials. The cubic R Br net  work is stabilized by either P or As, new interstitials for rare-earth-metal halide chemistry, while the heavier Sb and Bi become part of the R Pn host with Br as the interstitial   in hexagonal (and presumably metallic) R Pn Br phases.   At first, this might be viewed as a result of only a change in size of Z, but the two types of compounds formed are

175 0022-4596/99 $30.00 Copyright  1999 by Academic Press All rights of reproduction in any form reserved.

176

JENSEN, HOISTAD, AND CORBETT

fundamentally different. Though both involve some degree of R—R bonding, the cluster halides are more salt-like while the R Pn Z members are more intermetallic, so we are   comparing the addition of R—Pn bonding in a R—Br condensed cluster host in the first type with new R—Br bonding in a stable R Pn framework in the second.   2. EXPERIMENTAL

2.1. Syntheses All reagents and products were stored and handled in a nitrogen-filled glove box because of their sensitivity to oxygen and moisture. The rare-earth metals came from Ames Lab (599.99% purity), and any surface films (oxides or ‘‘tarnish’’) were scraped off before use. Phosphorus (Aldrich, 5—9’s), As (Aldrich, 6—9’s), Sb (Alfa, 5—9’s), and Bi (AESAR, 6—9’s) were used as received. LaBr and PrBr   were prepared from the oxides by the ammonium halide method (14) and purified by vacuum sublimation in contact with only Ta (15). Typically, &250 mg of reactants were loaded into Nb tubes that had previously been cleaned, crimped, and welded on one end. The open ends of the tubes were then tightly crimped, arc-welded under argon, and sealed into well-evacuated, baked fused silica jackets before being heated in the furnace. Phases were identified and yields were estimated visually from Guinier powder diffraction patterns of the products through routine comparisons with calculated patterns, as regard to both line positions and intensities. Powder patterns were also used to determine the lattice parameters of the new phases listed in Table 1 via least squares refinement of line positions relative to Si (NIST) added as an internal standard. During the initial explorations, many wide-ranging stoichiometries and reaction conditions were tried. ROBr was

TABLE 1 Lattice Constants (As ) and Cell Volumes (As 3) of Some R3Br3Pn and R5Pn3Br Phasesa,b Composition La Br P   La Br As   La Sb Br   La Bi Br   Pr Br P   Pr Br As   Pr Sb Br   Pr Bi Br  

Number of lines

a

c

»

42 38 36 47 22 12 40 38

11.970(2) 12.090(3) 9.602(2) 9.718(5) 11.746(8) 12.02(3) 9.378(2) 9.528(4)

— — 6.7331(3) 6.753(6) — — 6.529(4) 6.618(7)

1715.1(5) 1767.2(8) 537.6(2) 552.3(6) 1620(2) 1737(8) 497.3(3) 520.3(6)

?R Br Pn, I4 32, No. 214, Z"8; R Pn Br, P6 /mcm, No.193, Z"2.       @Dimensions from least-squares refinement of Guinier powder diffraction pattern with Si as an internal standard; j"1.54056 As , 23°C.

the most common unwanted product, appearing in almost every powder pattern, sometimes as the only impurity. This originated from adventitious water (and perhaps even O ), mostly as evolved from the fused silica on  heating. For La Br P, the best results were obtained from mix  tures loaded near this stoichiometry (yields 85—100%) and heated to 900—1000°C for 3—4 weeks followed by slow cooling to room temperature (10°C/h). Continued heating at 800—850°C for 4 weeks did not improve yields. Lower yields were obtained when other reactant ratios were tried, while the binary compounds LaP and LaBr appeared in  the powder patterns of the products at the extremes. The stuffed-Mn Si structure was never seen, even when the   La P Br stoichiometry was loaded.   The synthesis of the corresponding arsenide in 80—90% yields was accomplished similarly at 900—950°C. A small amount of AlBr in the reaction (as a transporting agent for  LaBr (16)) produced 85% yields at 850°C after 6 weeks.  Other stoichiometries and similar reaction conditions gave lower yields (5—75%) and other phases, commonly LaBr  and LaAs when well off composition. The hexagonal structure was never seen, even when the stoichiometry La As Br   was loaded. Some weak lines in several complicated patterns were unidentifiable. La Sb Br was perhaps the easiest phase of this series to   prepare. Even the stoichiometries that produced the cubic phase for Pn"P or As gave instead some of the hexagonal phase with Sb after 950—1000°C for 2—4 weeks. LaBr ,  LaOBr, and, less often, LaSb sometimes were found when proportions were varied. Yields of at least 90% could be achieved from the correct reactant proportions. Reactions to gain La Bi Br were loaded on stoichiometry and resulted   in 90% of the intended product plus 10% LaOBr after 2 weeks at 950°C. Other experiments targeted on La Sb M (M"Ti, V, Cr,   Mn, Fe, Co, Zr) with added KBr, AlBr , or LaBr as   potential fluxing or transport agents and run in tantalum at 900 or 950°C for 2 weeks produced La Sb Br in yields of   590% (experiments with AlBr ), 40—70% (LaBr ), or 10%   (KBr). In all but the AlBr trials, the impurities were mix tures of LaSb, LaSb , LaOBr, and KBr/LaBr . In some   AlBr and LaBr cases, 5—10% of the powder pattern lines   could not be assigned, and it is supposed that these products may have contained other phases incorporating the transition metals. The analogues Pr Br Z and Pr Z Br were loaded on     stoichiometry and heated at 950°C for 2 weeks. The yields were high: 90% Pr Br P, 80% Pr Br As, &100%     Pr Sb Br, and 90% Pr Bi Br, respectively, with the same     structure type distribution. The yield of Pr Br As was lower   than for any other phase prepared. PrOBr was the only identifiable impurity, but a few lines (5%) in two of the powder patterns could not be assigned.

177

LANTHANUM AND PRESEODYMIUM BROMIDE PNICTIDES

2.2. Crystallography Laue photographs were used to determine the suitability of possible crystals of these phases. Although the powder patterns showed large amounts of the target products, the only usable crystals came from a reaction that had been loaded for La Sb Cr with LaBr added. Later experiments    in which no Cr was present resulted in an identical powder pattern, so it was possible to identify the irregular black crystal as La Sb Br. A crystal was mounted in a thin-walled   glass capillary and sealed off, and diffraction data were collected at room temperature with graphite-monochromated MoKa radiation on a Rigaku AFC6R diffractometer. Routine indexing and cell reduction readily gave a primitive hexagonal cell. The 2027 reflections measured without condition up to 2h"50° led, after an absorption correction based on four t-scans, to 171 unique reflections (R "5.2%). Systematic absences indicated three possible  space groups: acentric P6 cm (No. 185) and P6c2 (No. 188)  and centric P6 /mcm (No. 193). The structure was solved by  direct methods (17) and refined in the expected centric group with the program TEXSAN (18). Isotropic refinements converged at R(F)"3.4% and R "4.0% and an isotropically at R"1.7% and R "2.0%. The largest  residual peak in the final difference map was 0.94 e/As . A summary of the crystallographic and refinement para-

TABLE 2 Summary of Data Collection and Refinement Parameters for La5Sb3Bra Empirical formula Formula weight Crystal size (mm); habit Space group,? Z D , g/cm   Diffractometer Radiation, j (As ) k, cm\ (MoKa) Temperature, °C Octants measured Scan mode 2h ,°

 Total reflections measured, R (%)  Observed reflections (I53p ) ' Unique reflections (I53p ) ' Variables Relative transmission coefficient range Residuals, R; R @  Goodness of fit (*o) , (*o) , e/As 



 Secondary ext. coeff. (10\)

La Sb Br   1139.7 0.1;0.2;0.2; black, irregular P6 /mcm (No. 193), 2  7.04 Rigaku AFC6R MoKa, 0.71069, graphite-monochromated 305.85 23 h,$k, $l u—2h 50 2027, 5.2 1467 171 14 0.763—1.000 1.7; 2.0 1.82 0.94, !2.1 8.7 (4)

?Lattice parameters in Table 1. @R" " " F "!" F " "/ " F .    R "[ w( " F "!" F ")/ w(F )]; w"[p(F)]\.    

TABLE 3 Positional and Thermal Parameters for La5Sb3Br x

Atom La1 La2 Sb Br

y

z B (As )? 

  0   0 0.2744(1)    0.3827(1) x  0 0 0

0.57(3) 0.60(6) 0.60(7) 0.84(7)

º @ 

º 

º 

0.0060(5) 0.0081(8) 0.0094(5) 0.0081(4) º 0.0056(6)  0.0068(6) 0.0080(8) 0.0083(6) 0.0112(8) º 0.010(1) 

?B "(8n/3) º a*a*a a .  G H GH G H G H @¹"exp[!2n(º ha*#º kb*#º lc*#2º hka*b*#2     º hla*c*#2º klb*c*)], º "º "º "0.     

meters are in Table 2, and atomic positional and displacement parameters are given in Table 3. Structure factor data are available from J.D.C. 3. RESULTS AND DISCUSSION

3.1. R Br Pn   Although no suitable single crystals of the cubic phase could be found, the powder diffraction line positions and interstices for the black La Br P, La Br As, Pr Br P, and       Pr Br As agreed very well with those calculated with the   refined lattice constants (Table 1) for the Ca I P structure  type with positional parameters as refined earlier for Pr I Pt (6). Trigonal antiprisms (‘‘octahedra’’) of the rare earth metal atoms centered by Pt, P, etc., share three nonadjacent edges with neighboring octahedra while the other nine edges are bridged by multiple-functioned bromine atoms. Figure 1 shows the repeat unit, with heavier lines representing shared R—R edges. This building block clearly derives from the well-known noncondensed R X Z-type   clusters. In the present cases, a P or As atom occupies the center of each octahedron. The sharing creates a three-

TABLE 4 Important Interatomic Distances (As ) and Angles (°) in La5Sb3Br La1—La1 La1—La2 La1—Sb La2—Sb La2—Br La2—La2 Sb—Sb Br—Br

3.3666(2) 3.8997(9) 3.4334(7) 3.281(1) 3.1269(9) 4.2753(7) 4.051(1) 3.3665(2)

La1—La1—La2 La2—La1—La2 La1—La2—La1 La1—La1—Sb La2—La1—Sb La1—La2—Sb

64.434(7) 66.48(2) 51.14(1) 60.648(7) 52.79(2)? 57.01(1)@ 56.30(1)? 54.81(1)@

La2—Br—La2 86.25(1)A 93.75(1)B Sb—La1—Sb 93.08(3) Sb—La2—Br 78.17(2)? La1—Sb—La2 70.85(2)? 68.18(1)@ La2—Sb—La2 86.25(1)? La1—Sb—La1 58.72(1)

?La2 and Sb at z"1/4. @La2 at z"1/4, Sb at z"!3/4. ALa atoms on same face of an octahedron with different z’s. BLa at same z. CDiagonal La atoms.

178

JENSEN, HOISTAD, AND CORBETT

FIG. 1. One octahedral (R Br Pn)Br unit in cubic R Br Pn, Pn"P      or As. Each octahedron shares the three heavier-lined nonadjacent edges not bridged by Br with neighboring octahedra (6).

dimensional network of helical 4 chains of clusters as  shown in Fig. 2, where circles represent clusters and ‘‘bonds’’ the shared edges (6). In contrast, the monoclinic Pr Br Ru   (7) and R I Z (19) phases with the same stoichiometry  consist of R I Z clusters that share both trans and some cis   R—R edges to generate double chains of clusters. As is usual, the condensed cluster networks in these cubic R X Z phases afford metallic compounds. This and the   bonding have been considered before (6). The way in which the size of the unit cell depends on the identities of the atoms is illustrated by the first two compounds in Table 1. With lanthanum, the cubic lattice dimensions increase by 0.12 As on changing the P interstitial to As, and the unit-cell volume increases 3%. The same substitution in the smaller Pr compounds results in an increase of 0.27 As in a while the unit-cell volume increases almost 7%.

3.2. R Pn Br   One might expect the substitution of Sb for As to result in an even larger cubic lattice constants. However, no examples of R Br Sb (or R Br Bi) could be found. Rather, the     hexagonal stuffed Mn Si -type structure results for these   R Pn Br. This ternary structure type for rare-earth metal   hosts has been reported earlier for La Ge Z (Z"B , C ,   V V

FIG. 2. The 4 helices (with one darkened) in the R Br Pn compounds.    Each sphere represents an octahedral unit as in Fig. 1, and the ‘‘bonds,’’ the shared edges.

N—Sb, O—Se, Cl, Cr—Zn, Ru, Cd) (11), a similarly large family of La Pb Z (12), and R Pn H (R"Eu,Yb) (13). The     present bromides represent the first halide analogues of the last hydride family. According to powder diffraction data studies, only the (La,Pr) (Sb,Bi) binaries have been re  ported before with presumably unadorned Mn Si -type   structures (9). These all contain infinite chains of confacial trigonal antiprisms (‘‘octahedra’’)  [R(2) Pn ] that run    along 0, 0, z together with 1D linear chains of R atoms,  [R (1) Pn ] along  ,  , z, etc., with a short R—R repeat of      c/2. Both chains involve common Pn. Figure 3 shows the refined result for La Sb Br. The chains of condensed anti  prismatic La are surrounded by one type of Sb atom with  three functions, bridging edges of the shared faces of the trigonal antiprisms, bonding exo to R2 in neighboring chains (not shown), and defining distorted trigonal antiprisms around R1 in the 1D metal strings. Bromine centers every La antiprism. The squat antiprism seen in Fig. 3 and the relatively small c/a value appear to be characteristic

LANTHANUM AND PRESEODYMIUM BROMIDE PNICTIDES

FIG. 3. A view perpendicular to the c axis of hexagonal La Sb Br   (P6 /mcm) showing the chain of confacial trigonal antiprisms of La2 atoms  as well as two parallel chains of La1 atoms. The black Sb atoms coordinate both types of gray La atoms while the interstitial Br atoms are shown open (90% probability thermal ellipsoids).

of period 5 main-group elements as they are similar to those found in Zr Sn Z (20) and Zr Sb Z (21) with     greater numbers of excess electrons than for R Pn X   (5 ) 3—3 ) 3—1"5). The binary La Sb has been reported twice before with   fairly consistent lattice constants, a"9.42(1), 9.428 As and c"6.62(1), 6.618 As (22, 23), a necessary but not sufficient condition to ensure that the common impurity Z (H, C, O, etc.) were not present. The addition of the interstitial bromine atom to these supposed binary La Sb examples in  creases both parameters uniformly by &1.8% and the volume by 5.5% (Table 1) as expected from studies of the known La Ge Z derivatives, considering that the larger   main-group element Sb is present. On the other hand, the consistent literature data for Pr Sb (22, 24) afford less   believable increases on bromide addition, 1.6% in a but only 0.3% (0.019(6) As ) in c, especially when it is noted that the added bromine Z atoms are still only 3.26 As (c/2) apart along that axis. (Carbon is known to produce such a contraction in c in La Ge and La Pb (11,12).) Furthermore,    

179

comparisons of the present dimensions of the bismuth bromide phases with literature data for R Bi require more   careful consideration because there is good evidence that at least La Bi exists only as a ternary compound (i.e.,   a Nowotny phase (10)). First, two sets of literature data for the supposed binary La Bi (25, 26) differ by 0.045(2) As   (0.5%) in a, but only 0.003 As (1 p) in c, good evidence of the presence of interstitial impurities in at least one instance. The larger of these sets of dimensions yields a 2.1% volume increase on incorporation of Br. More importantly, we have been unable to obtain Mn Si -type La Bi after reaction of     the pure elements sealed in a tantalum container with heating in an induction furnace under vacuum (to remove H), only the reported anti-Th P -type La Bi appeared (27).     Similar studies of Pr Bi have not been found, and it is more   difficult to judge whether apparent increases of cell lengths and volume of 0.8, 1.0, and 2.6%, respectively, are reasonable. The presence of the large main-group element Bi probably soaks up some of these changes, as it does for the analogous La Pb Z (12).   An important question regarding these results is why the basic structure type changes so radically with such fairly small changes in the pnictide, i.e., from the cubic La Br Pn   for P and As to hexagonal La Pn Br with Sb and Bi. Size   parallels offer the simplest explanation, especially as far as the known hexagonal R Pn binary systems. On this basis,   P and As may be too small relative to lanthanum, even before introduction of the large bromine interstitial. A check of Pearson’s Handbook (9) reveals that none of the 17 R Pn compounds listed with the Mn Si structure type     contains P, and only two contain As, Eu As and Yb As ,     both with presumably divalent R. On the other hand, all (but PrAs) of the possible RPn that we saw frequently as alternate products have been reported in the cubic NaCl structure, which is close to the Ca I P structure but without  the interstitial. The implication is that the smaller pnictogens do not form the hexagonal and more intermetallic structure type, with or without an interstitial halogen, particularly with the large R. Evidence regarding the instability of cubic R Br Pn for   Pn"Sb, Bi is more difficult to find. Perhaps there is only the obvious reason, that La Pn Br (or other) analogues are   more stable. The cubic structure is known for (La,Pr) X Z   for X"Br or I and generally for the later 4d and 5d interstitials (6, 7). Examples of Gd Cl Z, Z"C, Si are also   stable (5). Other prospective combinations have simply not been tested, but this structure is generally not wideranging, judging from the results of diverse explorations. Significantly, no rare-earth metal cluster or condensed cluster halide has been found with interstitial elements anywhere near Sb and Bi, and the present P and As are the first reports with these. (The closest were formerly Cu and Si (2).) The elements Sb and Bi would not be expected to generate very strong R—Z bonding in these relatively polar

180

JENSEN, HOISTAD, AND CORBETT

systems, as contrasted with the more intermetallic R Pn   hosts. Finally, it should be noted that although a great many R Pn Br compositions were explored in the eight V W X systems studied here, only one other new ternary phase was encountered, a new layered lanthanum bromide phosphide (28).

ACKNOWLEDGMENTS This research was supported by the National Science Foundation, Solid State Chemistry, via Grant DMR-9510278 and was carried out in the facilities of the Ames Laboratory, U.S. Department of Energy.

REFERENCES 1. A. Simon, Hj. Mattausch, G. J. Miller, W. Bauhofer, and R. K. Kremer, in ‘‘Handbook on the Physics and Chemistry of Rare Earths’’ (K. A. Gschneidner and L. Eyring, Eds.), Vol. 15, pp. 191—285. North-Holland. Amsterdam, 1991. 2. J. D. Corbett, J. Alloys Compd. 229, 10 (1985). 3. S. Uma and J. D. Corbett, Inorg. Chem. 37, 1944 (1998). 4. J. D. Corbett, J. Chem Soc. Dalton ¹rans. 575 (1996). 5. E. Warkentin and A. Simon, Rev. Chim. Miner. 20, 488 (1983). 6. P. K. Dorhout, M. W. Payne, and J. D. Corbett, Inorg. Chem. 30, 4960 (1991). 7. R. Llusar and J. D. Corbett, Inorg. Chem. 33, 849 (1994). 8. C. Hamon, R. Marchand, Y. Laurent, and J. Lang, Bull. Soc. Fr. Miner. Crystallogr. 97, 6 (1974).

9. P. Villars and L. D. Calvert, ‘‘Pearson’s Handbook of Crystallographic Data for Intermetallic Phases,’’ 2nd ed. American Society for Metals International, Materials Park, OH, 1991. 10. J. D. Corbett, E. Garcia, A. M. Guloy, W.-M. Hurng, Y.-U. Kwon, and A. E. Leon-Escamilla, Chem. Mater. 10, 2824 (1998). 11. A. M. Guloy and J. D. Corbett, Inorg. Chem. 32, 3532 (1993). 12. A. M. Guloy and J. D. Corbett, J. Solid State Chem. 109, 352 (1994). 13. E. A. Leon-Escamilla and J. D. Corbett, J. Alloy Compd. 265, 104 (1998). 14. G. Meyer, S. Do¨tsch, and T. Staffel, J. ¸ess-Common Met. 127, 155 (1987). 15. J. D. Corbett, Inorg. Synth. 22, 31 (1983). 16. G. N. Papatheodorou, in ‘‘Current Topics in Material Science’’ (E. Kaldis, Ed.), Vol. 10, p. 258, North-Holland, New York, 1982. 17. G. M. Sheldrick, SHELXS-86, Programs for Structure Determination, Universita¨t Go¨ttingen, Germany, 1986. 18. TEXSAN, Version 6.0, Molecular Structure Corp., The Woodlands, TX, 1990. 19. M. W. Payne, P. K. Dorhout, S.-J. Kim, T. R. Hughbanks, and J. D. Corbett, Inorg. Chem. 31, 1389 (1992). 20. Y.-U. Kwon and J. D. Corbett, Chem. Mater. 4, 1348 (1992). 21. E. Garcia and J. D. Corbett, Inorg. Chem. 29, 3274 (1990). 22. W. Rieger and E. Parthe´, Acta Crystallogr., Sect. B 24, 456 (1968). 23. G. Borzone, A. Borsese, A. Saccone, and R. Ferro, J. ¸ess-Common Met. 65, 253 (1979). 24. M. N. Abdusalyanova, O. I. Rahmaton, N. D. Faslyeva, and A. G. Tehuiko, J. ¸ess-Common Met. 141, L23 (1988). 25. K. Yoshihara, J. B. Taylor, L. D. Calvert, and J. G. Despault, J. ¸ess-Common Met. 41, 329 (1975). 26. K. Nomura, H. Hayakawa, and S. Ono, J. ¸ess-Common Met. 52, 259 (1977). 27. E. A. Leon-Escamilla, Ph.D. dissertation, Iowa State University, 1996. 28. E. A. Jensen and J. D. Corbett, unpublished research.